Elucidating the factors that control the replication fidelity of DNA polymerases is a goal of great fundamental and practical importance. In order to advance in this direction we will generate realistic atomic-level simulations of the catalysis and replication fidelity of wild-type and mutant DNA polymerases, focusing on DNA polymerase (pol), which is a key player In human base excision repair and has been Implicated in the incidence of cancer. The main strength of our laboratories has been our combined expertise in state-of- the art simulations of enzyme catalysis and catalytic landscapes, with a deep understanding of the available experimental information on the structural and mechanistic underpinnings of the fidelity of DNA polymerases. Our theoretical toolbox includes the empirical valence bond (EVB), the paradynamics that provides effective way of obtaining ab initio quantum mechanical /molecular mechanics (QM/MM) free energy profiles using the EVB as a reference potential, as well as coarse grained (CG) renormalizatlon approaches for exploring long time coupling between the conformational and chemical coordinates. We have also develop and refined effective sampling methods that should allow us to increase the accuracy of the calculated free energies. These computational studies will be applied in concert with the biochemical studies of Project 3 of kinetic effects of mutations of pol and of changes in its substrates. Our simulations will reproduce and/or predict the functional effects of these changes and analyze their origin. At the same time, we will rely on the important structural information from Project 1. The atomic level understanding Of the structure of the transition state for the insertion of a new nucleotide by pol and the malleability of this structure by modifications of the active site will greatly increase the possibilities of effective design of a potent and selective inhibitor of pol .